Skip to content

Latest commit

 

History

History
251 lines (160 loc) · 6.13 KB

File metadata and controls

251 lines (160 loc) · 6.13 KB

Simplified AI Assistant Project - Chat Interface Developer Roles

Dev 1: Backend (FastAPI + LangChain integration) Dev 2: Frontend (Streamlit Chat UI) + GitHub Setup Dev 3: Prompt Engineering + Documentation + Conversation Flow

Day 1: Setup & Foundation Dev 1 (Backend):

Set up FastAPI project structure Install dependencies (fastapi, langchain, uvicorn) Create basic API with /chat endpoint for conversation

Dev 2 (Frontend + GitHub):

Create GitHub repo with branches (main, dev, prompt-design, ui, api-integration) Set up Kanban board on GitHub Projects Create Streamlit project with st.chat_message and st.chat_input

Dev 3 (Prompts + Docs):

Design conversational flow for each feature:

Bot asks questions to gather info User responds naturally

Create initial system prompts for chat assistant Define conversation stages (greeting → questions → generation)

Day 2: Core Development Dev 1 (Backend):

Implement LangChain chat setup with memory/context Create /chat endpoint that:

Receives user messages Maintains conversation history Returns AI responses

Add session management for conversations

Dev 2 (Frontend):

Build chat interface with Streamlit:

Message history display Chat input box User/Assistant message styling

Implement conversation state management Add "Start New Conversation" button

Dev 3 (Prompts + Flow):

Design conversational prompts:

Bio flow: "Hi! I'll help you create a bio. What's your name?" → "What are your key skills?" → etc. Project flow: "Let's create a project summary! What's the project name?" → "What technologies did you use?" → etc. Reflection flow: "I'll help you write a learning reflection. What topic did you learn about?" → etc.

Define when bot should ask follow-up questions vs. generate final output

Day 3: Integration & Conversation Logic Dev 1 (Backend):

Integrate Dev 3's conversational prompts Add logic to detect when to:

Ask follow-up questions Generate final output Offer to start new topic

Implement conversation context tracking

Dev 2 (Frontend):

Connect chat UI to backend /chat endpoint Display typing indicators during API calls Add buttons for quick actions:

"Generate Bio" "Summarize Project" "Write Reflection"

Implement copy-to-clipboard for final outputs

Dev 3 (Integration):

Work with Dev 1 on conversation state management Test conversation flows end-to-end Refine prompts based on testing Create conversation scripts/examples

Day 4: Polish & Features Dev 1 (Backend):

Add ability to regenerate responses Implement conversation export functionality Add error handling for failed responses Optimize response time

Dev 2 (Frontend):

Add sidebar with:

Conversation history Quick action buttons Examples section

Improve chat UI styling Add message timestamps Implement scroll-to-bottom on new messages

Dev 3 (UX + Docs):

Add welcome message with instructions Create example conversations for each feature Add helpful hints in chat (e.g., "Tip: Be specific about your skills") Write user guide documentation Test conversation flows with edge cases

Day 5: Final Polish & Demo Dev 1 (Backend):

Final testing and optimization Code cleanup Merge branches

Dev 2 (Frontend):

Final UI polish and styling Add branding/logo Ensure responsive design Prepare deployment

Dev 3 (Documentation):

Complete README.md with:

Setup instructions How to use the chat interface Example conversations

Document prompt engineering approach Create demo video/presentation Final end-to-end testing

Jira Board - 15 Cards Epic: Project Setup (Day 1)

[Dev 2] Create GitHub repo, branches, and Kanban board [Dev 1] Set up FastAPI with /chat endpoint structure [Dev 2] Build basic Streamlit chat interface [Dev 3] Design conversational flows for all 3 features

Epic: Backend Development (Day 2-3)

[Dev 1] Implement LangChain chat with conversation memory [Dev 1] Add session management and context tracking [Dev 1] Integrate conversational prompts and generation logic

Epic: Frontend Development (Day 2-3)

[Dev 2] Build chat UI with message history and input [Dev 2] Add quick action buttons and navigation [Dev 2] Connect frontend to backend chat API

Epic: Conversation Design (Day 2-3)

[Dev 3] Create system prompts for bio, summary, and reflection [Dev 3] Test and refine conversation flows

Epic: Polish & Features (Day 4)

[Dev 1 + Dev 2] Add regenerate, export, and UI enhancements [Dev 3] Add examples, welcome message, and user guidance

Epic: Documentation & Demo (Day 5)

[Dev 3] Complete documentation and demo preparation

Chat Flow Example Bio Generation Conversation: 🤖 Assistant: Hi! I'm your AI assistant. I can help you create: • Professional bios • Project summaries
• Learning reflections

What would you like to create today?

👤 User: I need a professional bio

🤖 Assistant: Great! I'll help you create a compelling bio. Let's start with your name - what should I call you?

👤 User: John Doe

🤖 Assistant: Nice to meet you, John! What are your main skills or areas of expertise?

👤 User: Python, machine learning, data analysis

🤖 Assistant: Excellent! Can you share 2-3 key achievements or projects you're proud of?

👤 User: Built a recommendation system, led a team of 5 devs

🤖 Assistant: Perfect! What tone would you like - professional, casual, or creative?

👤 User: Professional

🤖 Assistant: Here's your professional bio:

[Generated Bio Content]

Would you like me to regenerate it, or shall we work on something else?

Key Features of Chat Interface ✅ Natural conversation - feels like chatting with a human assistant ✅ Context-aware - remembers what user said earlier ✅ Guided questions - bot asks for missing information ✅ Quick actions - buttons to jump to specific features ✅ Copy & export - easy to save generated content ✅ Regenerate option - can refine outputs ✅ Conversation history - see past chats in sidebar

Daily Coordination Day 2 Evening: Dev 3 shares conversation scripts with Dev 1 Day 3 Morning: Dev 1 confirms chat API ready for Dev 2 Day 3 Afternoon: All devs test conversation flows together Day 4 Morning: Team reviews UX and makes improvements

This chat-based approach makes the app more engaging and user-friendly! 🚀